Rent a private office. Preferably with 24-hour access. Wifi is a bonus and commonly included. It helps to have dedicated workspaces. A coffee shop could work if it has a good setup for coding & good coffee. Create the space for the project, and then you'll have the time for side projects.
As for design it takes time but being able to dedicate time to the top-down grand scheme and then go bottom-up is crucial for any size project. What are the fundamental requirements? Keep it simple. Deconstruct all aspects into single-purpose functions. One input, one output, & repeat. The ongoing pattern is to change the level of abstraction used for reviewing the project at any range, from a function to the whole project. These reflexes take time to build, and working on a side project daily is an excellent way to discover what patterns work best for your programming style and the project you're working on.
And big project ideas provide a field of side projects to discover. So wade into the water and find some fundamental piece of your larger idea. And then zoom in on that. When you're considering optimizations of compiled code to run a single binary on a static empty container image, you've gone deep enough.
Indeed, a reasonable cause for this is the latency in the cloud disk maintenance process. A cascade of disk failures in the same hour, and backups are spun up, along with other integrity checks, is a possible source of the vanishing file event.
I generate unit tests, comments, and more. A critical aspect is well-defined system instructions. Defining system instructions and prompts with XML formatting provides significant benefits. Think of LLMs as a force multiplier for most use cases. Upload the codebase and generate comments, unit tests, functional tests, mc/dc tests, etc. What follows is a matter of review and edits to outputs. Never unquestioningly trust, and instead, slowly train and refine models for any given project, as this approach tends to reap rewards in my experience.
Looks good & like fun so far. Ruff, mypy, and pytest could be helpful. And if you got that far I prefer nuitka for optimizing the python as a standaloned single file binary.
Study the area and the various tribes in their natural habitat. Study the different forms of communication and reciprocity conducted by such people. M. Sahlins categorized Generalized, Balanced, and Negative reciprocity. How are these exchanges performed, and in what context do these activities occur? How do you optimize your goals while adequately meeting the implicit and explicit socio-cultural norms of the tribes you seek to form bonds with? Building a living example of your skills and an interactive history via projects imbued with the information gathered from your studies is an essential task before total immersion with these tribes.
Thank you! That's my applied learning from Richard M. Eastman's book Style: Writing and reading as the discovery of outlook, 3rd edition. My blog and primary developer site are currently under active development. My github is @https://github.com/nortosem
Updates about the blog and links will appear there. The open-source projects are on standby while I finish rolling up my custom stack and ui/ux for the dev site.
Another way to solve this is to use Fossil. Fossil has a built-in wiki. You can launch Fossil UI and use the built-in wiki with Markdown. Have legacy txt files? Just open a new fossil project and add them to it. And if you set up the admin and user correctly, you can mirror your notes to GitHub. So, it's not that text files are not a good option. There are vastly superior options with almost no effort.
Another way to solve this is to use Fossil. Fossil has a built-in wiki. You can launch Fossil UI and use the built-in wiki with Markdown. Have legacy txt files? Just open a new fossil project and add them to it. And if you set up the admin and user correctly, you can mirror your notes to GitHub. So, it's not that text files are not a good option. There are many vastly superior options with almost no effort.
Spreadsheet programs such as LibreOffice are the next level. These are the most advanced and easily customized text files yet. Think of these as multidimensional text files that are all connected in an endless grid. Text files may seem ok, but managing tens of thousands of pages across sheets and books seems more straightforward with the spreadsheet format. But Vim is a great fallback when sheets are overkill.
As for design it takes time but being able to dedicate time to the top-down grand scheme and then go bottom-up is crucial for any size project. What are the fundamental requirements? Keep it simple. Deconstruct all aspects into single-purpose functions. One input, one output, & repeat. The ongoing pattern is to change the level of abstraction used for reviewing the project at any range, from a function to the whole project. These reflexes take time to build, and working on a side project daily is an excellent way to discover what patterns work best for your programming style and the project you're working on.
And big project ideas provide a field of side projects to discover. So wade into the water and find some fundamental piece of your larger idea. And then zoom in on that. When you're considering optimizations of compiled code to run a single binary on a static empty container image, you've gone deep enough.